Grammomys Surdaster, the Natural Host for Plasmodium Berghei Parasites, As a Model to Study Whole-Organism Vaccines Against Malaria

Grammomys surdaster, the Natural Host for Plasmodium berghei Parasites, as a Model to Study Whole-Organism Vaccines against Malaria

Solomon Conteh,1 Charles Anderson,1 Lynn Lambert,1 Sachy Orr-Gonzalez,1 Jessica Herrod,1 Yvette L. Robbins,1 Dariyen Carter,1 Stomy Bin Shamamba Karhemere,2 Pati Pyana,2 Philippe Büscher,3 and Patrick E. Duffy1* 1Laboratory of Malaria Immunology and Vaccinology (LMIV), National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH), Rockville, Maryland; 2Department of Parasitology, Institut National de Recherche Biomédicale (INRB), Kinshasa, Democratic Republic of Congo; 3Department of Biomedical Sciences, Unit of Parasite Diagnostics, Institute of Tropical Medicine, Antwerp, Belgium

Abstract. Inbred mice are commonly used to test candidate malaria vaccines, but have been unreliable for predicting efficacy in humans. To establish a more rigorous animal model, we acquired African woodland thicket rats of the genus Grammomys, the natural hosts for Plasmodium berghei. Thicket rats were acquired and identified as Grammomys surdaster by skull and teeth measurements and mitochondrial DNA genotyping. Herein, we demonstrate that thicket rats are highly susceptible to infection by P. berghei, and moderately susceptible to Plasmodium yoelii and Plasmodium chabaudi:1–2 infected mosquito bites or 25–100 sporozoites administered by intravenous injection consistently resulted in patent parasitemia with P. berghei, and resulted in patent parasitemia with P. yoelii and P. chabaudi strains for at least 50% of animals. We then assessed efficacy of whole-organism vaccines to induce sterile immunity, and compared the thicket rat model to conventional mouse models. Using P. berghei ANKA radiation-attenuated sporozoites, and P. berghei ANKA and P. yoelii chemoprophylaxis vaccination approaches, we found that standard doses of vaccine sufficient to protect laboratory mice for a long duration against malaria challenge, are insufficient to protect thicket rats, which require higher doses of vaccine to achieve even short-term sterile immunity. Thicket rats may offer a more stringent and pertinent model for evaluating whole-organism vaccines.

INTRODUCTION tory mice and therefore more pertinent to humans than are laboratory mice. After decades of malaria vaccine development, a highly efficacious vaccine to prevent infection remains a top prior- ity. Inbred mice have long been the animal model most MATERIALS AND METHODS commonly used to test candidate vaccines, yet have failed Rodent parasites. to predict outcomes in humans. For example, BALB/c mice Plasmodium berghei ANKA and Py develop sterile immunity against Plasmodium berghei (Pb) 17XNL stabilates (cryopreserved, frozen, blood-stage par- infection after a single immunization with 1,000 Pb ANKA asites) were stored at the National Institutes of Health radiation-attenuated sporozoites (RAS),1 and 90–100% pro- (NIH) Laboratory of Malaria Immunology and Vaccinology tection against Plasmodium yoelii (Py) with very few PyRAS (LMIV), the Pb NK65 stabilate was obtained from the NIH (750 given intravenously × 3 doses).2 In comparison, humans Laboratory of Malaria and Vector Research (LMVR), and that received much higher doses and numbers of immuni- the Pcc AS and Pcc CB stabilates were obtained from zations have achieved inconsistent levels of protection.3–6 Jean Langhorne, Division of Parasitology, Medical Research Consideration of this difference led Chatterjee and others Council National Institute for Medical Research, London, to investigate Pb in African woodland thicket rats (TRs) or United Kingdom. TRs and mice. Grammomys surdaster, a natural host–parasite combination. African woodland TRs were originally In this model, infection evolves in a chronic manner, similar captured in the wild in the south of the Democratic Repub- to human infection and in contrast to infection in laboratory lic of Congo, Africa, obtained from Department of Parasitol- mice that frequently results in high parasite loads. Their stud- ogy of Institut National de Recherche Biomédicale (INRB), ies suggested that TRs are highly susceptible to infection and bred under pathogen-free conditions at NIH/National and are more difficult to protect via RAS immunization, com- Institute of Allergy and Infectious Diseases (NIAID)/LMIV. parable to what is observed in humans with Plasmodium TRs are outbred and housed as couples or harem for life. 3,7 The offspring were weaned and removed from parents falciparum. Herein, we seek to establish a thriving colony 21–28 days postdelivery and housed separately. Laboratory of TR, confirm previous findings, and test whether TRs are mice (C57BL/6, BALB/c, and CD1) were obtained from NIH- susceptible to other prominent strains of rodent malaria, approved vendors. All breeding procedures and experiments including Py and Plasmodium chabaudi chabaudi (Pcc) were done in accordance with LMIV 1E ASP approved by strains. We also characterize whether RAS immunization the Institutional Animal Care and Use Committees. or chemoprophylaxis vaccination (CVac) confer protection Identification of TRs. African woodland TRs were identi- against various rodent malaria strains in TR. Finally, we con- fied using four specimen heads and other body parts. DNA firm that TRs are a more stringent animal model than labora- was extracted from biopsies taken from an external ear pinna from each, and 1,141 base pairs of mitochondrial DNA sequence were determined for each sample as previ- *Address correspondence to Patrick Duffy, National Institutes ously described.8 Further identification was made by stan- of Health, National Institute of Allergy and Infectious Diseases, Laboratory of Malaria Immunology and Vaccinology, 5640 Fishers dard measurements of skulls and teeth, and by direct visual Lane, TWB1/1111, Rockville, MD 20852. E-mail: patrick.duffy@ comparison with study skins and skulls in the National nih.gov Museum of Natural History (NMNH). 835 836 CONTEH AND OTHERS

Mosquitoes. Three- to ten-day-old female Anopheles vary glands were recorded as administering an infectious stephensi mosquitoes were obtained from the LMVR. The bite. The percent of infected mosquitoes used or the preva- mosquitoes were infected by allowing them to feed on a lence of infection of the mosquitoes used for the experi- donor mouse (BALB/c, CD1, or C57BL/6 mouse) during ments ranged from 70% to 90%, depending on the batch. blood-stage parasitemia. The donor mouse selected for the RAS immunization and homologous challenge. Using feed exhibited gametocyte counts of 1+ or 2+ (one or two PbRAS, groups of 10 C57BL/6 mice and 10 TRs were gametocytes per field for up to 10 fields) and a parasitemia immunized three times with 104 PbRAS at 3- to 4-week level of less than 10%. The infected mosquitoes were intervals, and another group of 10 TRs was immunized reared at 24°C and 80% relative humidity for the Py strain, three times total with 105 PbRAS for the first dose, and 5 × 26°C and 80% relative humidity for the Pcc strains, and 104 PbRAS for the second and third doses. Groups of five 19–21°C and 80% relative humidity for the Pb strain. The naïve C57Bl/6 mice and five TRs were used as controls. Four infected mosquitoes were maintained on water containing to five weeks after the last immunization, groups received 10% Karo syrup. Sporozoites (SPZ) were ready to use from an intravenous homologous challenge of 1,000 PbSPZ. 14 to 21 days post blood meal (dpb) for Py and Pcc strains, CVac and challenge. Using the nonirradiated SPZ con- and from 18 to 25 dpb for the Pb strain. centration per milliliter, the vaccination dose under chemo- Radiation and dissection of infected mosquitoes. Infected prophylaxis (pyrimethamine [PYR]) was calculated. PYR mosquitoes were irradiated or attenuated with 10,000– drug was administered orally via acidified drinking water 15,000 rads (100–150 Gy) using a cesium irradiator. The irra- from 5 hours before nonirradiated SPZ injection until up to diated or nonirradiated mosquitoes were placed in a freezer 42 hours after nonirradiated SPZ injection. PYR drug water (−20°C) for 3–5 minutes to immobilize them. The immobilized was then replaced with normal acidified drinking water. The mosquitoes were washed in a petri dish containing 70% following PYR drug stock concentration was prepared: 14 mg ethanol for 1 minute and were dried using a Kim wipe of PYR/mL of dimethyl sulfoxide with the stock diluted 100 (Kimberly-Clark Europe Limited, Reigate, Surrey, United timesinacidifieddrinkingwater(1:100=3mLofstockPYR: Kingdom) and transferred to another petri dish containing 1× 297 mL of acidified drinking water). For intraperitoneal drug phosphate-buffered saline (PBS) or E199. Mosquitoes were treatment, PYR stock concentration was diluted 100 times dissected at room temperature and salivary glands were (1:100 = 100 of stock PYR: 9900 of PBS) and 200 μLof removed using a dissecting microscope (Zeiss Stemi 2000-C diluted PYR was administered intraperitoneally at the fol- [Edmund Optics, Barrington, NJ] or Olympus SZ61 [Tokyo, lowing time points: 20, 27, and 42 hours after nonirradiated Japan]), 3/10-mL insulin syringes (U10029G1/2), and forceps. SPZ injection or vaccination. After each vaccination, blood One hundred pairs of dissected glands were put into a Protein smears were collected on day 7 and 14 to check for blood- LoBind 1.5-mL Eppendorf tube (Hamburg, Germany) containing stage infections. Four to five weeks or 16 weeks after the 100 μL of dissecting medium, 5% normal mouse serum in last immunization, groups received an intravenous homolo- PBS or E199 medium. After dissection, 0.2 mL of dissection gous challenge of 200 PySPZ. medium was added to increase the total volume to 0.3 mL. A 1-mL syringe with a 26G needle attached was used to tritu- RESULTS rate and release the SPZ from the glands. The total volume of processed salivary gland was increased to 0.5 or 1.0 mL, TRs are identified as G. surdaster. To establish a more depending on the expected SPZ yield per mosquito. SPZ rigorous animal model for testing malaria vaccines, we were counted using a Phase contrast microscope (Nikon acquired African woodland TRs of the genus Grammomys ECLIPSE E200 [Tokyo, Japan] or Zeiss Axiostar Plus (Figure 1), the natural hosts for Pb. TRs were transferred [Gottingen, Germany]) and a disposable hemocytometer. from a colony at INRB in Kinshasa to establish a breeding Ten microliters of medium containing SPZ per sample were colony at NIAID in Rockville, MD. TRs used in this study loaded in the hemocytometer and the SPZ concentration per were originally captured in the wild at Fungurume (200 km milliliter was determined. from Lubumbashi) and Lumata (50 km from Lubumbashi) in Infectious dose titrations in TR. Using the nonirradiated SPZ concentration per milliliter, the highest infectious dose was calculated and serial dilutions (1:2) were made to titrate the amount of SPZ needed per TR (400, 200, 100, 50, and 25). TR, C57Bl/6, and BALB/c mice were infected with each dose titra- tion using Pb ANKA SPZ, Pb NK65 SPZ, Py 17XNL SPZ, Pcc CB, and Pcc AS. Each infectious SPZ dose was administered in 200-μL dose volume intravenously via the tail vein. Daily blood smears were prepared on a slide and stained with 10% Giemsa to determine blood-stage parasitemia levels by microscopy. Each TR was exposed to 5–8 infected mosquitoes at a time, and mosquitoes were replaced only once if they did not probe or feed. Each mosquito pint was fitted with a FIGURE 1. Photos of thicket rat (TR; Grammomys surdaster ). net screen through which the mosquitoes probed or bit for Color photos of African woodland TR originally captured in the wild a blood meal. The test mosquitoes in the pints were immo- in the south of Democratic Republic of Congo, Africa, obtained − from Department of Parasitology, Institut National de Recherche bilized in a 20°C freezer for 3 minutes and then examined Biomédicale, and bred under pathogen-free conditions at National under a dissecting microscope. The mosquitoes with blood Institutes of Health/National Institute of Allergy and Infectious meals were dissected, and only those with SPZ in their sali- Diseases/Laboratory of Malaria Immunology and Vaccinology. THICKET RAT MODEL FOR MALARIA VACCINE EVALUATION 837

FIGURE 2. Thicket rats (TRs) are susceptible to infection with two strains of Plasmodium Berghei (Pb). (A) Dose-dependent parasitemia in three mice and three TRs inoculated with 25, 50, or 200 Pb ANKA sporozoites (SPZ). TRs were treated with chloroquine on day 6 after SPZ inoculation. (B) Parasitemia in four C57BL/6 mice and four TRs inoculated with 25, 50, 100, or 200 Pb NK65 SPZ. Each line represents one animal. the Katanga Province, south of the Democratic Republic of TRs are susceptible to infection by rodent malaria Congo, and have been maintained in a breeding colony at species. To establish reliable sporozoite doses to cause INRB, before transfer and rearing for these studies at LMIV. parasitemia, we first tested the susceptibility of TR to infec- The areas where TRs were originally captured correspond tion by parasites used in rodent malaria models. As Pb is with the regions where Pb strains were first isolated from commonly used in C57BL/6 mice, we dissected Pb ANKA G. surdaster and Anopheles dureni.9,10 and Pb NK65 salivary gland SPZ and compared their infec- Animals were identified as G. surdaster by mitochondrial tivity in TR to that in C57BL/6 mice. Initial infections deter- DNA genotyping and by skull and teeth measurements. All mined that Pb ANKA was lethal to TR, and subsequent mitochondrial sequences from four independent animals were infection experiments required chloroquine cure once the identical. When blasted to NCBI nonredundant protein data- blood-stage infection was imminent. Using dose titrations base (https://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastp), of between 25 and 200 SPZ injected intravenously, as few the top (best) hit was to accession number AF141218.1, for cyto- as 25 Pb ANKA (Figure 2A) or Pb NK65 SPZ (Figure 2B) chrome B from a Grammomys species from Tanzania with resulted in patency; mice and TRs appeared to have similar 93% identity. The second best hit was to accession number susceptibility to Pb ANKA, whereas susceptibility to Pb EU349747.1 for G. surdaster with 86% identity. Compara- NK65 was as high or higher in TR than in mice. To assess tively, the percent identity to laboratory mice and laboratory infection by natural transmission, TRs were exposed to rats was 83%, with approximately 880 other NCBI protein A. stephensi mosquitoes infected with Pb ANKA. Four of four sequences of cytochrome B being more similar to TR than TRs that provided blood meals to one or two sporozoite- laboratory mice or rats. The methods used to identify our TR infected mosquitoes became infected (Table 1). Two out of are previously described.11 Further identification was made by the three TR that did not provide blood meals to sporozoite- standard measurements of skulls and teeth taken with digital infected mosquitoes, nevertheless became infected pre- calipers accurate to 0.02 mm and compared with those sumably because they were probed by sporozoite-infected in Table 2 of Dieterlen and others12 and the measurements mosquitoes that did not take a blood meal. As few as one listedinHappoldandothers.13 Species determinations were infected bite was sufficient to induce parasitemia. made by direct visual comparison with study skins and skulls in the NMNH. The skulls best match Grammomys dolichurus among the species represented in the NMNH collection, which is synonymous with G. surdaster by current taxonomy. This identification is consistent with current understanding of species distributions and definitions within Grammomys.12,13

TABLE 1 TR infection with Plasmodium berghei ANKA by mosquito bite Fed mosquitoes TR Fed mosquitoes with sporozoites Parasitemia Experiment 1 1 5 1 Positive 2 3 2 Positive 3 0 0* Positive Experiment 2 1 3 1 Positive FIGURE 3. Thicket rats (TRs) are susceptible to infection with 2 2 2 Positive Plasmodium yoelii (Py) 17XNL. Parasitemia in four BALB/c mice 3 0 0* Positive and four TRs inoculated with 50, 100, 200, or 400 Py 17XNL sporo- 4 1 0 Negative zoites (SPZ). Mice self-cured by day 16 postinfection, whereas TR TR = thicket rat. also self-cured by day 23 postinfection. Each line represents one *Denotes mosquitoes that likely probed without taking a blood meal. animal. 838 CONTEH AND OTHERS

TABLE 2 per dose) at 3- to 4-week intervals. In contrast, this regimen TR infection with Plasmodium yoelii 17XNL by mosquito bite protected 0% of TRs against challenge with 1,000 PbSPZ. 5 Fed mosquitoes A higher immunizing dose of 2 × 10 PbRAS administered TR Fed mosquitoes with sporozoites Parasitemia in three doses at 3- to 4-week intervals (105 +5×104 +5× 1 2 1 Positive 104) protected 20% of TRs from challenge with 1,000 2 5 3 Negative 3 3 1 Positive PbSPZ (Figure 5). 4 3 3 Negative We next tested whole-sporozoite immunization of TRs TR = thicket rat. against Py using nonattenuated SPZ administered under drug cover. Because we wanted to limit parasite development to the liver, we used PYR, which has been shown to Next, Py 17XNL was used to infect TR and BALB/c mice. be a successful approach with Pb in C57BL/6 and BALB/c Salivary gland SPZ were dissected and injected intra- mice.14 Using a CVac approach, we again observed differ- venously in a dose range of 50–400 SPZ per animal. TRs ences between laboratory mice and TRs. After three doses were easily infected by Py, though not with the same sensi- of 3,000 Py 17XNL SPZ (PySPZ) under PYR drug cover at tivity as seen with Pb, with 100 or more Py SPZ resulting in 3- to 4-week intervals, 75% of BALB/c mice versus 0% patency in both animal models (Figure 3). Interestingly, while TRs were protected from challenge with 200 PySPZ mice and TR self-cured, TR appeared to require ∼1 week (Table 4). After three doses of 104 PySPZ under PYR treat- longer to clear parasitemia when SPZ were injected intrave- ment at 3- to 4-week intervals, 100% of mice were nously. We also confirmed infection by the natural route; protected but only 30% of TRs were protected from chal- two of four TRs that provided blood meals to sporozoite- lenge with 200 PySPZ 5 weeks postvaccination (Figure 6A). infected mosquitoes became infected. As few as two mos- When challenge with 200 PySPZ was delivered 15 weeks quito bites were sufficient to result in patency (Table 2). postvaccination, 90% of the mice were protected and Finally, we confirmed TR were susceptible to Pcc CB none of the TRs were protected (Figure 6B). (Figure 4A, Supplemental Figure 1) and AS (Figure 4B, Sup- plemental Figure 2) by both intravenous injection and mos- DISCUSSION quito bite (Table 3). Two of four TRs that provided blood meals to sporozoite-infected mosquitoes became infected, Laboratory mice infected with rodent Plasmodium spe- and as few as three infectious mosquito bites were suffi- cies such as P. berghei, P. yoelii, and P. chabaudi have cient to result in patency. In each mosquito bite experiment, been commonly used as proxies to investigate malaria all mosquitoes that probed took a blood meal except for in vivo. These models have been essential in developing one feeding study in which Pb ANKA-infected mosquitoes an understanding of the mechanisms of infection in a mam- probed without taking a blood meal; in this study, TRs were malian host, the progression of the disease, host immune not fully anesthetized and thus not completely still during responses, and parasite genetics.15 Rodent infection mosquito exposure, thus preventing the mosquitoes from models have also been instrumental in drug and vaccine taking a blood meal. development. However, many study results have been diffi- Whole-sporozoite vaccines have greater efficacy in cult to translate into clinical applications in humans, owing laboratory mice than TRs. We assessed the potential of in part to the physiological and genetic differences of TRs as a model to study whole-organism vaccines. C57BL/ the host and parasite species used. In particular, vaccine 6 mice were 100% protected (10/10) against challenge with studies in current mouse models of malaria have not been 1,000 PbSPZ (Figure 5) after immunization with a total of reliably predictive of human efficacy; laboratory mice are 3×104 PbRAS administered in three doses (104 PbRAS generally much easier to protect than humans. Although

FIGURE 4. Thicket rats (TRs) are susceptible to infection with Plasmodium chabaudi CB and AS. (A) Parasitemia in four TRs inoculated with 50, 100, 200, or 400 Plasmodium chabaudi chabaudi (Pcc) CB sporozoites (SPZ). (B) Parasitemia in four TRs inoculated with 100, 200, 400, or 800 Pcc AS SPZ. Each line represents one animal. THICKET RAT MODEL FOR MALARIA VACCINE EVALUATION 839

TABLE 3 TABLE 4 TR infection with Plasmodium chabaudi chabaudi CB by mosquito Plasmodium yoelii 17XNL pyrimethamine chemoprophylaxis vacci- bite nation against 200 P. yoelii SPZ challenge Fed mosquitoes 3×103 SPZ × 3 TR Fed mosquitoes with sporozoites Parasitemia No. protected/no. challenged % Protected 1 3 2 Negative 2 3 2 Positive TR 0/4 0 3 6 2 Negative BALB/c 3/4 75 4 5 3 Positive TR = thicket rat; SPZ = sporozoite. TR = thicket rat. these models have been good tools to screen out ineffec- mental models for Pb infection, characterizing the course of tive vaccines, they have not been as successful in selecting infection in wild rats as chronically mild with low parasitemia, effective vaccines.16 A more rigorous rodent model that is in contrast to the severe, fulminant disease observed in not easily protected via immunization would be useful to laboratory-bred species.22,23 Yoeli’s studies also provided qualify or prioritize potential vaccine candidates. early insight to the histologic pathology in liver tissue induced A new phylogenetic analysis from the Sackler Institute for by Pb sporozoite inoculation in T. surdaster.24 Subsequent to Comparative Genomics at the American Museum of Natural those early studies, rodents formerly known as T. surdaster History shows that malarial parasites found in tree-dwelling are now referred to as G. surdaster.25 rats share a sister relationship with P. falciparum and The TRs transferred from Kinshasa for these studies were Plasmodium reichenowi.17 Although Plasmodium parasites identified by a team of experts from the Laboratory of in rodents differ from Plasmodium parasites in humans Zoonotic Pathogens, Rocky Mountain Laboratories, and the (e.g., the existence of the PfEMP1 antigen family in NMNH Smithsonian Institute to be G. dolichurus, which is P. falciparum but not in rodent parasites18,19), individual currently regarded as a synonym for G. surdaster. The gene analysis, regardless of methods, consistently shows name likely represents a species composite in need of that P. falciparum and P. reichenowi aresistertaxaand taxonomic revision. Such a qualifier would simply acknowl- the three species that infect rodents were always recov- edge the appreciable increase in recognized rodent species ered as a monophyletic group.20,21 Many of the analyses that is occurring concurrent with the increased application placed P. falciparum and P. reichenowi as sisters to of molecular tools to taxonomic problems.26 In the future, P. chabaudi, P. berghei,andP. yoelii as described previ- TRs in the northern part of the Democratic Republic of Congo ously by Perkins.17 Since the malarial parasites found in may be separated out and referred to as G. surdaster, rodents share a sister relationship with P. falciparum,devel- whereas the TRs in the south of Democratic Republic of oping a TR model to study malaria may specifically acceler- Congo (LMIV TR) might continue to be referred to as ate progress toward the goal of a human malaria vaccine G. dolichurus. The area where these TRs were captured is the against P. falciparum. same area where Pb was originally isolated,9,10 hence the Early studies by Yoeli in the 1960s and 1970s investi- evidence favors that this is the natural host of Pb. gated the use of “tree rats” (Thamnomys surdaster)asexperi- Our results support the findings of Chatterjee and others,7 which suggested that TR (G. surdaster), the natural host for Pb, would be a more stringent rodent model to extrapolate to human malaria. TRs are highly susceptible to infection with doses as low as 25 Pb SPZ, and develop parasitemia after as few as one infected mosquito bite. We expand on previous findings by establishing TR susceptibility to other rodent forms of malaria including Py and Pcc, which resulted in patent parasitemia in at least 50% of animals. Confirming previous findings, the natural host was more difficult to protect by RAS immunization in contrast to artificial hosts like inbred mice.7 Similar to humans who require immunization with > 1,000 mosquitoes carrying PfRAS to achieve sterile immunity,27 TRs required higher doses of RAS to achieve only partial protection against homologous challenge. We expanded on previous findings by testing CVac methods of immunization, and observed similar trends: higher doses of CVac using PYR were required for protection in some TRs versus all inbred mice. Similar SPZ doses were used in both the RAS and CVac studies, FIGURE 5. Radiation-attenuated sporozoites (RAS) immunization fully protects C57BL/6 mice, but elicits partial protection in thicket whereas RAS experiments were conducted using Pb and rat (TR) against high-dose homologous Plasmodium berghei (Pb) CVac studies were conducted using Py. Our results confirm challenge. Percent parasite negative in C57BL/6 mice (N = 10) and that TRs are a more stringent rodent model for evaluating 4 – TR (N = 10) immunized three times with 10 RAS 3 4 weeks apart, whole-organism vaccines for malaria and, as seen in TR (N = 10) immunized three times with 105,5×104, and 5 × 104 RAS 3–4 weeks apart, and naïve controls (N = 5 per group) chal- humans, vaccine efficacy can be enhanced by increasing lenged with 1,000 Pb ANKA SPZ. vaccine dose.28 840 CONTEH AND OTHERS

FIGURE 6. High-dose chemoprophylaxis vaccination fully protects BALB/c mice but only partially protects thicket rat (TR) against homologous Plasmodium yoelii (Py) challenge. (A) Percent parasite negative in BALB/c and TR immunized three times via 104 Py sporozoites (SPZ) inoculation coupled with pyrimethamine (PYR) treatment and controls against challenge with 200 PySPZ 5 weeks postimmunization. (B) Percent parasite negative in BALB/c and TR immunized three times via 104 PySPZ inoculation coupled with PYR treatment and controls against challenge with 200 PySPZ 15 weeks postimmunization.

The different requirements to achieve protection in TRs REFERENCES and artificial models may be due to different mechanisms of protection; for example, innate immunity characterized by 1. Jaffe RI, Lowell GH, Gordon DM, 1990. Differences in susceptibility among mouse strains to infection with Plasmodium the infiltration of mononuclear cells, neutrophils, and eosin- berghei (ANKA clone) sporozoites and its relationship to ophils is prevalent in artificial hosts, whereas it is absent in protection by gamma-irradiated sporozoites. Am J Trop Med TRs.3,7 In addition, there may be differences in antigenic Hyg 42: 309–313. targets between the two models; for example, in laboratory 2. Chattopadhyay R, Conteh S, Li M, James ER, Epstein JE, mice, both immunizing and challenge SPZ were halted at Hoffman SL, 2009. The effects of radiation on the safety and protective efficacy of an attenuated Plasmodium yoelii the uninucleate, trophozoite stage, whereas development sporozoite malaria vaccine. Vaccine 27: 3675–3680. 3,7 to submature schizonts was observed in Pb in TRs. Per- 3. Chatterjee S, Perignon JL, Van Marck E, Druilhe P, 2006. How haps “the ‘adaptation’ of a parasite to an artificial host actually reliable are models for malaria vaccine development? Lessons translates in immunological terms into defense mechanisms from irradiated sporozoite immunizations. J Postgrad Med 52: 321–324. that in most circumstances are more effective than those 3 4. Bray RS, 1976. Vaccination against Plasmodium falciparum: seen in the natural host.” a negative result. Trans R Soc Trop Med Hyg 70: 258. 5. Edelman R, Hoffman SL, Davis JR, Beier M, Sztein MB, Received September 12, 2016. Accepted for publication November Losonsky G, Herrington DA, Eddy HA, Hollingdale MR, Gordon 29, 2016. DM, Clyde DF, 1993. Long-term persistence of sterile immunity in a volunteer immunized with X-irradiated Plasmodium Published online January 23, 2017. falciparum sporozoites. J Infect Dis 168: 1066–1070. Note: Supplemental figures appear at www.ajtmh.org. 6. Herrington D, Davis J, Nardin E, Beier M, Cortese J, Eddy H, Losonsky G, Hollingdale M, Sztein M, Levine M, Acknowledgments: We would like to thank Andre Laughinghouse Nussenzweig M, Clyde RS, Edelman DR, 1991. Successful and Kevin Lee of Laboratory of Malaria and Vector Research for immunization of humans with irradiated malaria sporozoites: providing all of the mosquitoes used in this project; Jean Langhorne humoral and cellular responses of the protected individuals. for providing the Plasmodium chabaudi chabaudi parasites; Robert Am J Trop Med Hyg 45: 539–547. Morrison for helping with the thicket rat (TR) sequence analysis; Tom 7. Chatterjee S, Ngonseu E, Van Overmeir C, Correwyn A, Druilhe G. Schwan from the Laboratory of Zoonotic Pathogens in Montana, P, Wery M, 2001. Rodent malaria in the natural host–irradiated and Darrin Lunde and Michael Carleton from the Smithsonian for sporozoites of Plasmodium berghei induce liver-stage specific their work to identify TR; and J. Patrick Gorres for help with writing immune responses in the natural host Grammomys surdaster and submitting the manuscript. and protect immunized Grammomys against P. berghei – Financial support: This research was supported and funded by the sporozoite challenge. Afr J Med Med Sci 30 (Suppl): 25 33. Intramural Research Program of the National Institute for Allergy 8. Lecompte E, Granjon L, Peterhans JK, Denys C, 2002. Cyto- and Infectious Disease of the National Institutes of Health. chrome b-based phylogeny of the Praomys group (Rodentia, Murinae): a new African radiation? C R Biol 325: 827–840. Authors’ addresses: Solomon Conteh, Charles Anderson, Lynn 9. Center LUM. Plasmodium berghei: Life Histories and Lambert, Sachy Orr-Gonzalez, Jessica Herrod, Yvette L. Robbins, Stabilates (Deep-Frozen Samples) of Isolates, Lines, and Dariyen Carter, and Patrick E. Duffy, Laboratory of Malaria Immu- Clones Maintained at the University of Edinburgh. Available nology and Vaccinology (LMIV), National Institute of Allergy and Infec- at: https://www.lumc.nl/sub/1040/att/811071156182556/ tious Diseases (NIAID), National Institutes of Health (NIH), Rockville, 811130228382556/908051142222557.pdf. Accessed Novem- MD, E-mails: [email protected], [email protected], ber 8, 2016. [email protected], [email protected], [email protected], yvette. 10. Killick-Kendrick R, 1974. Parasitic protozoa of the blood of [email protected], [email protected], and [email protected]. rodents: a revision of Plasmodium berghei. Parasitology 69: Stomy Bin Shamamba Karhemere and Pati Pyana, Department of 225–237. Parasitology, Institut National de Recherche Biomédicale (INRB), 11. Schwan TG, Anderson JM, Lopez JE, Fischer RJ, Raffel SJ, Kinshasa, Democratic Republic of Congo, E-mails: stomy_karhem@ McCoy BN, Safronetz D, Sogoba N, Maiga O, Traore SF, yahoo.fr and [email protected]. Philippe Büscher, Department of 2012. Endemic foci of the tick-borne relapsing fever spiro- Biomedical Sciences, Prince Leopold Institute of Tropical Medicine, chete Borrelia crocidurae in Mali, west Africa, and the poten- Antwerp, Belgium, E-mail: [email protected]. tial for human infection. PLoS Negl Trop Dis 6: e1924. THICKET RAT MODEL FOR MALARIA VACCINE EVALUATION 841

12. Dieterlen F, 2013. Genus Grammomys thicket rats. Happold 22. Yoeli M, Alger N, Most H, 1963. Tree rat, Thamnomys surdaster DCD, ed. Mammals of Africa, Vol. III. London, United Kingdom: surdaster, in laboratory research. Science 142: 1585–1586. Bloomsbury Publishing, 404–405. 23. Yoeli M, Most H, 1964. A study of Plasmodium berghei in 13. Happold DCD, 2013. Grammomys dolichurus woodland thicket Thamnomys surdaster, and in other experimental hosts. Am rat (Common Grammomys). Happold DCD, ed. Mammals of J Trop Med Hyg 13: 659–663. Africa, Vol. III. London, United Kingdom: Bloomsbury Pub- 24. Yoeli M, 1974. Slow maturing primary exo-erythrocytic schizonts lishing, 410–411. of Plasmodium berghei in an experimentally infected tree-rat 14. Matuschewski K, Hafalla JC, Borrmann S, Friesen J, 2011. (Thamnomys surdaster). Trans R Soc Trop Med Hyg 68: Arrested Plasmodium liver stages as experimental anti- 302–305. malaria vaccines. Hum Vaccin 7 (Suppl): 16–21. 25. Buscher P, Bin Shamamba SK, Ngoyi DM, Pyana P, Baelmans 15. Siu E, Ploss A, 2015. Modeling malaria in humanized mice: R, Magnus E, Van Overmeir C, 2005. Susceptibility of – opportunities and challenges. Ann N Y Acad Sci 1342: 29 36. Grammomys surdaster thicket rats to Trypanosoma brucei 16. Waters H, 2011. Better animal models needed for malaria gambiense infection. Trop Med Int Health 10: 850–855. vaccine development, experts say. Available at: http:// 26. Musser GG, Carlteon MD, 2005. Superfamily Muroidea. Wilson blogs.nature.com/spoonful/2011/10/better_animal_models_ D, Reeder D, eds. Mammal Species of the World: A Taxo- needed_fo_1.htm. Accessed. nomic and Geographic Reference, 3rd edition. Baltimore, 17. Perkins SL, 2008. Molecular systematics of the three mito- MD: Johns Hopkins University Press, 894–1531. chondrial protein-coding genes of malaria parasites: corrob- 27. Hoffman SL, Goh LM, Luke TC, Schneider I, Le TP, Doolan DL, orative and new evidence for the origins of human malaria. Sacci J, de la Vega P, Dowler M, Paul C, Gordon DM, Stoute Mitochondrial DNA 19: 471–478. JA, Church LW, Sedegah M, Heppner DG, Ballou WR, Richie 18. De Niz M, Ullrich AK, Heiber A, Blancke Soares A, Pick C, TL, 2002. Protection of humans against malaria by immuni- Lyck R, Keller D, Kaiser G, Prado M, Flemming S, Del zation with radiation-attenuated Plasmodium falciparum Portillo H, Janse CJ, Heussler V, Spielmann T, 2016. The – machinery underlying malaria parasite virulence is con- sporozoites. J Infect Dis 185: 1155 1164. served between rodent and human malaria parasites. Nat 28. Seder RA, Chang LJ, Enama ME, Zephir KL, Sarwar UN, Commun 7: 11659. Gordon IJ, Holman LA, James ER, Billingsley PF, Gunasekera 19. Larremore DB, Sundararaman SA, Liu W, Proto WR, Clauset A, A, Richman A, Chakravarty S, Manoj A, Velmurugan S, Li M, Loy DE, Speede S, Plenderleith LJ, Sharp PM, Hahn BH, Ruben AJ, Li T, Eappen AG, Stafford RE, Plummer SH, Rayner JC, Buckee CO, 2015. Ape parasite origins of human Hendel CS, Novik L, Costner PJ, Mendoza FH, Saunders malaria virulence genes. Nat Commun 6: 8368. JG, Nason MC, Richardson JH, Murphy J, Davidson SA, 20. Perkins SL, Sarkar IN, Carter R, 2007. The phylogeny of rodent Richie TL, Sedegah M, Sutamihardja A, Fahle GA, Lyke malaria parasites: simultaneous analysis across three genomes. KE, Laurens MB, Roederer M, Tewari K, Epstein JE, Sim Infect Genet Evol 7: 74–83. BK, Ledgerwood JE, Graham BS, Hoffman SL, Team VRCS, 21. Perkins SL, Schall JJ, 2002. A molecular phylogeny of malarial 2013. Protection against malaria by intravenous immuniza- parasites recovered from cytochrome b gene sequences. tion with a nonreplicating sporozoite vaccine. Science 341: J Parasitol 88: 972–978. 1359–1365.